Professional Documents
Culture Documents
Aspiration Pneumonia
Aspiration Pneumonia
Aspiration Pneumonia
Review Article
Aspiration Pneumonia
Lionel A. Mandell, M.D., and Michael S. Niederman, M.D.
A
spiration pneumonia is best considered not as a distinct entity From McMaster University, Hamilton,
but as part of a continuum that also includes community- and hospital- ON, Canada (L.A.M.); and Weill Cornell
Medical College, New York (M.S.N.). Ad-
acquired pneumonias. It is estimated that aspiration pneumonia accounts dress reprint requests to Dr. Mandell at
for 5 to 15% of cases of community-acquired pneumonia, but figures for hospital- lmandell@mcmaster.ca.
acquired pneumonia are unavailable.1 Robust diagnostic criteria for aspiration N Engl J Med 2019;380:651-63.
pneumonia are lacking, and as a result, studies of this disorder include heteroge- DOI: 10.1056/NEJMra1714562
neous patient populations. Copyright © 2019 Massachusetts Medical Society.
assumptions of lung sterility and of bacterial ac- such as glucocorticoids, have been shown in vi-
cess to the lungs through aspiration (microaspi- tro to promote the growth of S. pneumoniae and
ration or macroaspiration) and inhalation. Spe- certain gram-negative rods.13-16
cifically, genomic methods have defined a One hypothesis linking the airway microbiome
complex taxonomic landscape of bacteria in the with aspiration pneumonia is that illness may
lung and revealed the presence of diverse com- result in a change in the lung microbiota (dys-
munities of microbiota. We are still learning biosis), which may, in turn, interfere with or
about the role of the microbiome in health and impair pulmonary defenses. A macroaspiration
disease, as well as in the pathogenesis of pneu- event, particularly in a patient with risk factors
monia.5,6 In the healthy state, the immune tone for impaired bacterial elimination, such as re-
of the airways and alveoli appears to be calibrated duced consciousness or an impaired cough reflex,
by the bacteria constituting the lung microbiota, could then overwhelm the elimination side of the
an observation recently reported in both healthy immigration–elimination balance, further dis-
humans and experimental animal models.7,8 Con- rupting bacterial homeostasis and triggering an
cepts of virulence have also changed; virulence increase in a positive feedback loop leading to
is defined as “the relative capacity of a microor- acute infection.
ganism to cause damage to a host.”9 Infection is Bacteria may colonize various sites in the hu-
not simply the result of bacterial replication or of man oral cavity, such as the gingiva, dental plaque,
bacterial gene products; it is also a consequence and tongue.17,18 Pathogenic bacteria, including
of the host response, resultant inflammation, gram-negative species that are not seen in the
and tissue damage.9,10 normal host, may emerge in the elderly, as well
Models proposed to explain the possible role as in patients in nursing homes or hospitals and
of the lung microbiome in pneumonia include those with nasogastric tubes.19-21 Cleavage of cell-
the adapted island model of lung biogeography, surface fibronectin exposes receptors for gram-
effects of environmental gradients on lung micro- negative rods on underlying airway epithelial cells
biota (e.g., regional differences in oxygen ten- and is more closely correlated with host factors
sion and nutrient availability in the lungs), and such as acute illness than with the site of care.22
finally, pneumonia as an emerging phenomenon In the 1970s, anaerobes with or without aerobes
in the complex adaptive system of the lung micro- were the predominant pathogens in aspiration
bial ecosystem.11,12 The stability of the lung mi- pneumonia.23-26 More recently, there has been a
crobiome is probably maintained by a balance of shift to bacteria usually associated with commu-
immigration and elimination of bacteria and by nity- and hospital-acquired pneumonias, and
feedback loops. Immigration involves bacterial anaerobes are recovered less frequently.26 One
movement from the oropharynx to the lung pri- study of aspiration pneumonia in patients in the
marily by means of microaspiration, and elimi- intensive care unit showed that in community-
nation mainly occurs through ciliary clearance acquired cases, the main isolates were S. pneu-
and coughing. Negative and positive feedback moniae, Staphylococcus aureus, Haemophilus influenzae,
loops can suppress or magnify signals, respec- and Enterobacteriaceae, whereas gram-negative
tively, such as those for bacterial growth. An bacilli, including P. aeruginosa, were found with-
inflammatory event may lead to epithelial and out anaerobes in hospital-acquired cases.27 An-
endothelial injury, creating a positive feedback other study assessed the incidence of anaerobic
loop that can promote inflammation, disrupt bacteria in patients with ventilator-associated
bacterial homeostasis, and increase susceptibil- pneumonia and in those with aspiration pneu-
ity to infection. The complex adaptive system monia.28 Bacterial pneumonia was diagnosed in
model suggests that acute bacterial pneumonia 63 patients with ventilator-associated pneumonia
results from enhancement of a growth-promoting and in 12 patients with aspiration pneumonia.
signal by a positive feedback loop. This may re- Among patients with aspiration pneumonia, en-
sult in a rapid shift from a diverse microbial teric gram-negative organisms were isolated in the
mixture to dominance by a single species (e.g., patients with gastrointestinal disorders, whereas
Streptococcus pneumoniae or Pseudomonas aeruginosa).12 S. pneumoniae and H. influenzae predominated in
Various signaling molecules in humans, includ- those with community-acquired aspiration events.
ing neurotransmitters, cytokines, and hormones Only one anaerobic organism was found, and
the authors questioned the need for anaerobic tion at the time of extubation continue to have
coverage in both ventilator-associated pneumo- this problem at the time of discharge.33
nia and aspiration pneumonia.28 Additional risks include degenerative neuro-
Studies of the elderly continue to show the logic diseases (multiple sclerosis, parkinsonism,
trend away from anaerobes. A study involving 95 and dementia) and impaired consciousness, par-
institutionalized elderly patients with severe ticularly as a result of stroke and intracerebral
aspiration pneumonia reported 67 pathogens.29 hemorrhage, which can also impair cough clear-
Gram-negative enteric bacteria accounted for ance. The frequency of stroke-associated pneu-
49% of the pathogens, anaerobes for 16%, and monia is related to the severity of neurologic
S. aureus for 12%. Aerobic gram-negative bacteria illness and its associated immune impairment,
were found in conjunction with 55% of anaero- with higher rates among patients requiring inten-
bic isolates. Another study, involving 62 elderly sive care than among those admitted to a stroke
hospitalized patients with aspiration pneumonia, unit.34 Impaired consciousness can also result
showed that of 111 bacteria identified, gram- from drug overdose and medications, including
negative bacilli and anaerobes each accounted narcotic agents, general anesthetic agents, certain
for 19.8% of the bacteria, and anaerobes and antidepressant agents, and alcohol (Fig. 2A).
aerobes together were found in 66.7% of pa- After adjustment for other risk factors, antipsy-
tients who died.30 It is unclear why the patho- chotic medications increased the risk of aspira-
gens have changed, but it may be due to a shift tion pneumonia by a factor of 1.5 in a study in-
in the demographic characteristics of patients volving 146,552 hospitalized patients.35 Enteral
and earlier sampling today than in the past. feeding can lead to high-volume aspiration, espe-
Prior studies often collected cultures later in the cially when associated with gastric dysmotility,
illness, often after the development of empyema poor cough, and altered mental status. In three
or lung abscess.1 This discussion does not apply studies of enteral feeding after a stroke in a total
to chemical pneumonitis, which unlike aspiration of more than 5000 patients, early tube feeding
pneumonia, is a noninfectious, inflammatory improved survival, as compared with no feeding,
response of the airways and pulmonary paren- and in the first 2 to 3 weeks after the stroke,
chyma to acidic gastric contents or bile acids.1,31 nasogastric tube feeding was associated with
improved survival and functional outcomes, as
R isk Fac t or s compared with percutaneous enteral tube feed-
ing.36 Enteral feeding tubes are not currently
Aspiration is often the result of impaired swal- recommended for patients with dementia.37
lowing, which allows oral or gastric contents, Patients with multiple risks have increased
or both, to enter the lung, especially in patients rates of aspiration pneumonia, death, and other
who also have an ineffective cough reflex (Fig. 1). adverse outcomes. A meta-analysis of studies
Large-volume aspiration occurs with dysphagia; involving frail elderly patients showed that dys-
head, neck, and esophageal cancer; esophageal phagia increased the odds ratio for aspiration
stricture and motility disorders; chronic obstruc- pneumonia by a factor of 9.4, but when cerebro-
tive pulmonary disease; and seizures.1,3,32-41 In a vascular disease was also present, the odds ratio
case–control study involving elderly patients with rose to 12.9.38 In a study involving 1348 patients
pneumonia and healthy elderly controls, oropha- with community-acquired pneumonia, 13.8% of
ryngeal dysphagia increased the risk of pneumo- the patients were considered to be at risk for
nia (odds ratio, 11.9) and was present in nearly aspiration, and this subgroup of patients had a
92% of the patients who had pneumonia. Results higher 1-year mortality (hazard ratio, 1.73) and
of videofluoroscopic evaluation showed that increased risks of recurrent pneumonia (hazard
16.7% of the patients with pneumonia were able ratio, 3.13) and rehospitalization (hazard ratio,
to swallow safely, as compared with 80% of the 1.52) as compared with the rest of the study
controls.32 In extubated survivors of respiratory population.39 Similarly, a study involving 322
failure, dysphagia and aspiration are identified patients with community-acquired pneumonia
in at least 20% of patients. The frequency of identified the important risk factors for aspira-
swallowing dysfunction declines over time, but tion pneumonia as dementia (odds ratio, 5.20),
up to 35% of patients with swallowing dysfunc- poor performance status (odds ratio, 3.31), and
Risk Factors
Impaired swallowing
Esophageal disease: dysphagia, cancer, stricture
Chronic obstructive pulmonary disease
Brain–swallowing axis
Neurologic diseases: seizures, multiple sclerosis,
parkinsonism, stroke, dementia Oropharyngeal contents
Mechanical ventilation extubation Gastric contents
Impaired consciousness
Neurologic disease: stroke
Cardiac arrest
Medications
General anesthesia
Alcohol consumption
Increased chance of gastric contents reaching the lung
Reflux
Tube feeding
Impaired cough reflex
Medications
Alcohol
Stroke
Dementia
Degenerative neurologic disease
Impaired consciousness
use of sleeping pills (odds ratio, 2.08).40 Those pneumonia is cardiac arrest. In a study involving
with two or more risk factors had an increased 641 patients with cardiac arrest, pneumonia de-
incidence of recurrent pneumonia and increased veloped within 3 days after the event in 65% of
30-day and 6-month mortality, with rates rising the patients.41 The presumed mechanism is aspi-
in parallel with the number of risk factors. The ration of gastric contents during resuscitation
relationship between distinct risk factors for (promoted by stomach ventilation and the resus-
macroaspiration and the frequency and outcome citation procedure) and inhalation of oral secre-
of aspiration pneumonia underscores the differ- tions during bag–valve–mask ventilation and in-
ence between aspiration pneumonia and tradi- tubation. When therapeutic hypothermia to 33°C
tional community-acquired pneumonia: patients was used after cardiac arrest, the odds ratio for
with traditional community-acquired pneumo- early-onset pneumonia rose by a factor of 1.9.41
nia have no associated increase in the risk of However, a target temperature of 36°C may be
aspiration. associated with a lower risk of pneumonia.42
An important clinical context for aspiration Some studies showed that the incidence of early-
A B
C D
onset pneumonia decreased among patients re- Clinical features range from no symptoms to
ceiving systemic antibiotics at the time of car- severe distress with respiratory failure, and the
diac arrest.43 Two intervention studies involving clinical consequences may develop acutely, sub-
comatose patients showed a benefit of adminis- acutely, or slowly and progressively. Aspiration
tering prophylactic antibiotics for up to 24 hours into the lung can affect either the airway (caus-
after emergency intubation.44,45 ing bronchospasm, asthma, and chronic cough)
or the lung parenchyma. This discussion focuses
on aspiration into the lung parenchyma, which
Cl inic a l Fe at ur e s
can take the form of aspiration resulting in
Although macroaspiration is an essential feature chemical pneumonitis, aspiration of bland mate-
of aspiration pneumonia and chemical pneumo- rial (blood or the contents of tube feeding), or
nitis, many episodes are unwitnessed; therefore, aspiration resulting in bacterial pneumonia.
the magnitude of the exposure is often unknown. Aspiration pneumonia is usually acute, with
symptoms developing within hours to a few days and a pattern that is characteristic of acute respi-
after a sentinel event, although anaerobic aspira- ratory distress syndrome develops in up to 16.5%
tion may be subacute because of the less virulent of patients with witnessed aspiration, although
bacteria, and clinical features are difficult to the frequency rises if other risk factors (shock,
distinguish from those of other bacterial pneu- trauma, or pancreatitis) are also present 51
monias. In a study involving patients with pneu- (Fig. 2B). Low-pH aspirates are usually sterile, and
monia who were more than 80 years of age, bacterial infection is unusual initially, although
those with aspiration had a higher mortality, superinfection may develop subsequently.
higher serum sodium levels, and worse renal In most cases, neither chemical pneumonitis
function than patients without aspiration.46 In a nor aspiration pneumonia occurs with tube feed-
group of 53 patients with pneumonia and fluoro- ings or aspirated blood, since the aspirate pH is
scopically documented dysphagia, more patients usually high and uncontaminated by bacteria.
had bronchopneumonia than lobar pneumonia Although acid-suppressing therapy is associated
(68% vs. 15%), and 92% had posterior infiltrates.47 with an increased risk of community- or hospital-
Most patients with poor performance status had acquired pneumonia, which is related to gastric
diffuse and not focal infiltrates. Aspiration pneu- overgrowth by gram-negative bacteria, neutral-
monia is associated with higher mortality than ization of gastric pH may reduce the risk of
other forms of pneumonia acquired in the com- chemical pneumonitis.52 In a prospective cohort
munity (29.4% vs. 11.6%), a finding that may study involving 255 patients undergoing gastro-
have implications for hospitals that do not prop- intestinal endoscopy, the use of proton-pump
erly code its presence.48 In a survey of more than inhibitors or histamine H2 blockers was associ-
1 million patients in more than 4200 hospitals, ated with a significant reduction in the risk of
aspiration was documented in 4 to 26% of epi- gastric contents with a pH of <2.5 (odds ratio,
sodes of pneumonia. The expected mortality 0.24).53 Asphyxia may result if the aspirated vol-
among patients with aspiration pneumonia is ume of the bland material is large, but there may
higher than that for other forms of pneumonia, be few clinical findings if the volume is small.
and the risk-adjusted mortality (now used as a The chest radiograph may initially be abnormal,
quality metric) is lower for hospitals reporting until the aspirated material is cleared by suction
a high frequency of aspiration than for hospitals or coughing. In cases of unwitnessed aspiration,
reporting a low frequency of aspiration.48 it may be difficult to distinguish among chemi-
Macroaspiration of gastric contents can lead cal pneumonitis, aspiration pneumonia, and as-
to chemical pneumonitis but only with large- piration of bland material. An aspirated solid
volume, low-pH (usually <2.5) aspiration. In ani- foreign body can obstruct the airway and lead to
mal models, chemical pneumonitis develops only postobstructive pneumonia, further complicating
after exposure to at least 120 ml of gastric con- the distinction from bacterial pneumonia. In a
tents with a pH of 1. Described by Mendelson in series of patients with foreign-body aspiration
1946 as a consequence of obstetrical anesthesia, who were more than 65 years of age, the event
chemical pneumonitis is uncommon with mod- was recognized in only 29% of the patients,
ern anesthesia methods (1 case per 3216 proce- leading to a diagnostic delay of 1 to 3 months.54
dures), with a higher risk during emergency Chest radiographic findings were in the right
surgery and a lower risk with elective proce- lung in 65% of the patients, and food material
dures.49 In up to 64% of patients with aspiration accounted for more than 80% of the episodes.
during anesthesia, clinical or radiographic abnor-
malities do not develop. Lung injury from acid Di agnosis
aspiration is due to the release of inflammatory
mediators, including chemokines (e.g., interleu- The diagnosis of aspiration pneumonia depends
kin-8), proinflammatory cytokines (e.g., tumor on a characteristic clinical history (witnessed
necrosis factor), and neutrophil recruitment.50 macroaspiration), risk factors, and compatible
Chemical pneumonitis is characterized by a findings on chest radiography. These radiograph-
sudden onset of dyspnea, hypoxemia, tachycar- ic findings include infiltrates in gravity-dependent
dia, and diffuse wheezes or crackles on exami- lung segments (superior lower-lobe or posterior
nation. A chest radiograph is usually abnormal, upper-lobe segments, if the patient is in a supine
position during the event, or basal segments of with metronidazole failed in 5 of 11 cases of
the lower lobe, if the patient is upright during the lung abscess and in a randomized trial was less
event) (Fig. 2C and 2D). However, a chest radio- effective than clindamycin for anaerobic pulmo-
graph may be negative early in the course of nary infection.63,64
aspiration pneumonia. In a study involving 208 A prospective, randomized trial showed no
patients with pneumonia (more than 60% of significant differences among ampicillin–sulbac-
whom had aspiration), the chest radiograph was tam, clindamycin, and a carbapenem (panipenem–
negative in 28% of patients in whom pneumonia betamipron [not available in the United States])
was confirmed on computed tomography.55 One for the treatment of suspected aspiration pneu-
of the entities that can be confused with aspira- monia in elderly Japanese patients.65 A random-
tion pneumonia is negative-pressure pulmonary ized trial involving 96 patients compared moxi-
edema. It is important to consider this diagno- floxacin with ampicillin–sulbactam; both regimens
sis, which is accompanied by bilateral and gen- had clinical response rates of 66.7%.66 Anaerobes
erally symmetric lung infiltrates and is the result were isolated in 29.6% of patients with only lung
of breathing against a closed airway after general abscess. In those with only aspiration pneumo-
anesthesia, choking, or near drowning, all condi- nia, anaerobes were not found, and the most
tions that may also be accompanied by aspiration. frequent aerobes were Escherichia coli, Klebsiella
Although the diagnosis is usually clinical, pneumoniae, and P. aeruginosa (Fig. 2C).
some studies have used quantitative lung-lavage Antibiotic selection depends on the site of
cultures to distinguish bacterial from noninfec- acquisition (the community, a hospital, or a long-
tious (chemical and bland-aspiration) pneumo- term care facility) and risk factors for infection
nitis.56 Several investigations have studied bio- with multidrug-resistant pathogens. Risk factors
marker and biochemical measurements to predict include treatment with broad-spectrum antibiot-
bacterial infection after aspiration. A study involv- ics in the past 90 days and hospitalization for at
ing 65 intubated patients with risk factors for least 5 days. For most patients with community-
aspiration and a new lung infiltrate correlated acquired cases, treatment with ampicillin–sul-
quantitative bronchoalveolar-lavage cultures with bactam, a carbapenem (ertapenem), or a fluo-
serum procalcitonin levels.56 Measurement of roquinolone (levofloxacin or moxifloxacin) is
procalcitonin levels on days 1 and 3 did not dis- effective.3 In such patients, we suggest adding
tinguish the 32 patients with culture-positive clindamycin to another drug only when the risk
aspiration pneumonia from the 33 with culture- of predominantly anaerobic infection is high, as
negative pneumonitis. In studies of ventilated it is for patients with severe periodontal disease
patients, alpha-amylase levels (from salivary and and necrotizing pneumonia or lung abscess1
pancreatic sources) were elevated in airway secre- (Fig. 3 and Table 1). With mixed infection, elimi-
tions, at a frequency reflecting the number of risk nation of aerobic pathogens usually alters the
factors for aspiration, but the relevance of these local redox potential, eliminating anaerobes. For
findings to aspiration pneumonia and chemical hospital-acquired cases with a low risk of multi-
pneumonitis is not certain, and this is not a drug-resistant pathogens, a similar regimen may
method of value for diagnosis.57,58 be used. If resistance is a concern, broader-
spectrum treatment with piperacillin–tazobactam,
cefepime, levofloxacin, imipenem, or meropenem,
T r e atmen t
either singly or in combination, is required67
Aspiration Pneumonia (Fig. 3 and Table 1). In cases of multidrug-resis-
As documented pathogens have shifted from tant infection, an aminoglycoside or colistin may
anaerobes to aerobes, treatment regimens have be used as part of a combination regimen, with
also evolved. Even when anaerobes predominated the addition of vancomycin or linezolid if the
and penicillin was the drug of choice, penicillin- patient has documented nasal or respiratory
ase-producing anaerobes were reported.59,60 Com- colonization with methicillin-resistant S. aureus.
parative studies of anaerobic lung abscess and In a recent study involving comatose, mechan-
necrotizing pneumonia showed the superiority ically ventilated patients with aspiration, 43 of
of clindamycin, with penicillin failure attributed the 92 patients (46.7%) had bacterial aspiration
to resistant bacteroides species.61,62 Treatment pneumonia on the basis of bronchoscopic brush
Hospital-acquired or long-term
Community-acquired
care–acquired
If poor dental health, If normal dental If mild-to-moderate If severe illness If no risk of MDR, If there is a risk If mild-to-moderate If severe illness
nejm.org
(moxifloxacin), fluoroquinolone and MDR and (levofloxacin, carbapenem and MDR and
carbapenem (levofloxacin, reassess in 48–72 hr; moxifloxacin), (meropenem, reassess in 48–72 hr;
n e w e ng l a n d j o u r na l
For patients with suspected aspiration pneumonia, the decision about antibiotic therapy is dictated by the site of acquisition: community, hospital, or long-term care facility. Antibi-
otics are given to patients who have an aspiration event and an abnormal chest radiograph, although even with an initially normal radiograph, antibiotics are given to those with se-
vere illness (i.e., illness characterized by shock or requiring intubation). If chemical pneumonitis is suspected, antibiotics are not initially recommended, even with an abnormal ra-
diograph, unless the patient is severely ill; in patients with mild-to-moderate illness and an abnormal radiograph, the recommendation is to withhold antibiotics and reassess the
patient in 48 hours. Therapy is directed at the pathogens likely to be present in the patient at the time of aspiration, determined on the basis of the site of acquisition, with risk fac-
tors for resistant pathogens taken into consideration. Patients with severe illness are treated on the basis of risks associated with their dental health and multidrug resistance
Downloaded from nejm.org on February 13, 2019. For personal use only. No other uses without permission.
(MDR). Routine treatment for anaerobic pathogens is not needed in patients with normal dental health but should be considered in those with poor dental health (e.g., clindamycin
in patients with poor dental health and necrotizing pneumonia or lung abscess). If the site of acquisition is the community, outpatient treatment with amoxicillin–clavulanate, moxi-
floxacin, levofloxacin, or clindamycin can be administered orally. In the hospital, treatment is usually administered intravenously, but oral options are available for select patients; if
there is nasal or respiratory colonization with methicillin-resistant S. aureus, the addition of vancomycin or linezolid can be considered. BAL denotes bronchoalveolar lavage.
Aspir ation Pneumonia
samples.68 The investigators suggested that rou- Table 1. Antibiotic Treatment of Aspiration Pneumonia.*
tine antibiotic treatment should be started only if
bacterial infection is suspected but may be dis- Drug Dose, Schedule, and Route of Administration
continued if bronchoscopic cultures are negative. Ampicillin–sulbactam 1.5–3 g every 6 hr, intravenous
On the basis of data from studies of commu- Amoxicillin–clavulanate 875 mg twice daily, oral
nity-acquired pneumonia and hospital-acquired
Piperacillin–tazobactam 4.5 g every 8 hr or 3.375 g every 6 hr, intravenous
or ventilator-associated pneumonia, we suggest
Ceftriaxone 1–2 g once daily, intravenous
5 to 7 days of treatment for patients with a good
clinical response and no evidence of extrapulmo- Cefepime 2 g every 8–12 hr, intravenous
nary infection, and longer treatment for those Ertapenem 1 g once daily, intravenous
with necrotizing pneumonia, lung abscess, or em- Imipenem 500 mg every 6 hr or 1 g every 8 hr, intravenous
pyema. In the case of lung abscess or empyema, Meropenem 1 g every 8 hr, intravenous
drainage for diagnostic and treatment purposes Levofloxacin 750 mg once daily, intravenous or oral
may be needed. The choice of therapy should
Moxifloxacin 400 mg once daily, intravenous or oral
take into account potential drug-related adverse
events, including Clostridium difficile colitis and Clindamycin 450 mg three or four times daily, oral; or 600 mg
every 8 hr, intravenous
selection of antibiotic resistance. More data are
Gentamicin or tobra- 5–7 mg/kg once daily, intravenous
needed to identify the best antibiotic regimens mycin†
for aspiration pneumonia and to determine the
Amikacin† 15 mg/kg once daily, intravenous
duration of treatment. No randomized, controlled
Colistin‡ 9 million IU per day in two or three divided
trials have shown a role for glucocorticoids in doses, intravenous
the routine treatment of aspiration pneumonia,
Vancomycin† 15 mg/kg every 12 hr, intravenous
and we do not recommend their use.
Linezolid 600 mg every 12 hr, intravenous or oral
Chemical Pneumonitis
* Doses are for patients with normal renal function.
Initial treatment of gastric aspiration requires air- † The dose should be adjusted to a trough level of less than 1 mg per liter for
way maintenance, management of airway edema gentamicin and tobramycin, a trough level of less than 4 mg per liter for ami-
kacin, and a trough level of 10 to 15 µg per milliliter for vancomycin, with renal
or bronchospasm, and minimization of tissue function taken into consideration in all cases.
damage. Depending on the severity of the pneu- ‡ A loading dose of 6 million to 9 million IU can be administered.
monitis and the extent of care required, treatment
may include suctioning, bronchoscopy, intuba-
tion, mechanical ventilation, and intensive care. formed (Table 2). Medications known to pro-
Routine adjunctive treatment with glucocorticoids mote aspiration and interfere with swallowing
is not recommended, and antibiotics are not should be avoided, including sedatives, antipsy-
needed routinely unless the patient is taking acid- chotic agents, and for some at-risk patients, anti-
suppressing medication or has small-bowel ob- histamines.35 Aspiration-prevention efforts focus-
struction. In mild-to-moderate cases, we recom- ing on ventilator-associated pneumonia are not
mend withholding antibiotics even if there is discussed here.
radiographic evidence of an infiltrate, monitoring For patients with swallowing disorders, par-
clinical and radiographic findings, and reassess- ticularly after stroke, a full speech and swallow-
ing after 48 hours. In more serious cases, how- ing evaluation is necessary. Efforts should be
ever, antibiotics should be started empirically, and made to promote oral rather than enteral tube
the decision to continue antibiotic therapy for feeding, with the use of a mechanical soft diet
more than 2 to 3 days should be guided by the with thickened liquids rather than pureed food
clinical course. and thin liquids. In addition, “nutritional reha-
bilitation” with swallowing exercises and early
mobilization may help patients with dysphagia
Pr e v en t ion
and may prevent recurrence of aspiration pneu-
Postoperative chemical pneumonitis can be min- monia.69 Patients should receive enteral feeding
imized by ensuring that the patient has fasted in a semirecumbent rather than supine position
for at least 8 hours and has abstained from clear to minimize the risk of gastric aspiration. For
liquids for at least 2 hours before surgery is per- patients with oropharyngeal dysphagia, an effort
distinguish from other aspiration syndromes Dr. Mandell reports receiving fees for serving on a data moni-
toring committee from Bayer and Polyphor, fees for serving on a
and community- and hospital-acquired pneumo-
data and safety monitoring board from Cubist Pharmaceuticals
nias. The diagnosis should be considered in the and Shionogi, and lecture fees from Daiichi Sankyo; and Dr. Nieder-
appropriate clinical settings in patients with man, receiving consulting fees from Thermo Fisher, Paratek,
known risk factors for aspiration and character- Nabriva, Melinta, and Shionogi, advisory fees from Pfizer, and
grant support and consulting fees from Merck and Bayer. No other
istic clinical and radiographic findings. Aspira- potential conflict of interest relevant to this article was reported.
tion pneumonia is treated with antibiotics as Disclosure forms provided by the authors are available with
required but not with glucocorticoids. Preventive the full text of this article at NEJM.org.
We thank Dr. Robert P. Dickson, Division of Pulmonary and
measures should be used for patients at risk for Critical Care Medicine, University of Michigan, for his assistance
aspiration. and thoughtful comments on an earlier version of the manuscript.
References
1. Marik PE. Aspiration pneumonitis and Lung Cell Mol Physiol 2015;309:L1047- study of community- and hospital-acquired
aspiration pneumonia. N Engl J Med 2001; L1055. cases. Ann Intern Med 1974;81:329-31.
344:665-71. 14. Marks LR, Davidson BA, Knight PR, 26. Bartlett JG. How important are an-
2. Gleeson K, Eggli DF, Maxwell SL. Hakansson AP. Interkingdom signaling in- aerobic bacteria in aspiration pneumonia:
Quantitative aspiration during sleep in duces Streptococcus pneumoniae biofilm when should they be treated and what is
normal subjects. Chest 1997;111:1266-72. dispersion and transition from asymptom- optimal therapy. Infect Dis Clin North
3. DiBardino DM, Wunderink RG. Aspi- atic colonization to disease. MBio 2013; Am 2013;27:149-55.
ration pneumonia: a review of modern 4(4):e00438-13. 27. Mier L, Dreyfuss D, Darchy B, et al. Is
trends. J Crit Care 2015;30:40-8. 15. Belay T, Sonnenfeld G. Differential penicillin G an adequate initial treatment
4. Boaden E, Lyons M, Singhrao SK, et al. effects of catecholamines on in vitro for aspiration pneumonia? A prospective
Oral flora in acute stroke patients: a pro- growth of pathogenic bacteria. Life Sci evaluation using a protected specimen
spective exploratory observational study. 2002;71:447-56. brush and quantitative cultures. Intensive
Gerodontology 2017;34:343-56. 16. Freestone PP, Hirst RA, Sandrini SM, Care Med 1993;19:279-84.
5. Dickson RP, Erb-Downward JR, Huff- et al. Pseudomonas aeruginosa-catechola- 28. Marik PE, Careau P. The role of anaer-
nagle GB. The role of the bacterial micro- mine inotrope interactions: a contributory obes in patients with ventilator-associated
biome in lung disease. Expert Rev Respir factor in the development of ventilator- pneumonia and aspiration pneumonia:
Med 2013;7:245-57. associated pneumonia? Chest 2012;142: a prospective study. Chest 1999;115:178-83.
6. Segal LN, Rom WN, Weiden MD. 1200-10. 29. El-Solh AA, Pietrantoni C, Bhat A, et al.
Lung microbiome for clinicians: new dis- 17. Dewhirst FE, Chen T, Izard J, et al. Microbiology of severe aspiration pneu-
coveries about bugs in healthy and dis- The human oral microbiome. J Bacteriol monia in institutionalized elderly. Am J
eased lungs. Ann Am Thorac Soc 2014;11: 2010;192:5002-17. Respir Crit Care Med 2003;167:1650-4.
108-16. 18. Aas JA, Paster BJ, Stokes LN, Olsen I, 30. Tokuyasu H, Harada T, Watanabe E,
7. Segal LN, Clemente JC, Tsay JC, et al. Dewhirst FE. Defining the normal bacte- et al. Effectiveness of meropenem for the
Enrichment of the lung microbiome with rial flora of the oral cavity. J Clin Micro- treatment of aspiration pneumonia in el-
oral taxa is associated with lung inflam- biol 2005;43:5721-32. derly patients. Intern Med 2009;48:129-
mation of a Th17 phenotype. Nat Micro- 19. Kageyama S, Takeshita T, Furuta M, 35.
biol 2016;1:16031. et al. Relationships of variations in the 31. Wu YC, Hsu PK, Su KC, et al. Bile acid
8. Dickson RP, Erb-Downward JR, tongue microbiota and pneumonia mor- aspiration in suspected ventilator-associ-
Falkowski NR, Hunter EM, Ashley SL, tality in nursing home residents. J Geron- ated pneumonia. Chest 2009;136:118-24.
Huffnagle GB. The lung microbiota of tol A Biol Sci Med Sci 2018;73:1097-102. 32. Almirall J, Rofes L, Serra-Prat M, et al.
healthy mice are highly variable, cluster by 20. El-Solh AA, Pietrantoni C, Bhat A, et al. Oropharyngeal dysphagia is a risk factor
environment, and reflect variation in base- Colonization of dental plaques: a reser- for community-acquired pneumonia in the
line lung innate immunity. Am J Respir voir of respiratory pathogens for hospital- elderly. Eur Respir J 2013;41:923-8.
Crit Care Med 2018;198:497-508. acquired pneumonia in institutionalized 33. Macht M, White SD, Moss M. Swal-
9. Casadevall A, Pirofski LA. The damage- elders. Chest 2004;126:1575-82. lowing dysfunction after critical illness.
response framework of microbial patho- 21. Leibovitz A, Plotnikov G, Habot B, et al. Chest 2014;146:1681-9.
genesis. Nat Rev Microbiol 2003;1:17-24. Saliva secretion and oral flora in pro- 34. Hannawi Y, Hannawi B, Rao CP, Suarez
10. Casadevall A, Pirofski LA. Host- longed nasogastric tube-fed elderly pa- JI, Bershad EM. Stroke-associated pneu-
pathogen interactions: redefining the ba- tients. Isr Med Assoc J 2003;5:329-32. monia: major advances and obstacles.
sic concepts of virulence and pathogenic- 22. Johanson WG, Pierce AK, Sanford JP. Cerebrovasc Dis 2013;35:430-43.
ity. Infect Immun 1999;67:3703-13. Changing pharyngeal bacterial flora of 35. Herzig SJ, LaSalvia MT, Naidus E, et al.
11. Dickson RP, Erb-Downward JR, Mar- hospitalized patients — emergence of Antipsychotics and the risk of aspiration
tinez FJ, Huffnagle GB. The microbiome gram-negative bacilli. N Engl J Med 1969; pneumonia in individuals hospitalized for
and the respiratory tract. Annu Rev Physiol 281:1137-40. nonpsychiatric conditions: a cohort study.
2016;78:481-504. 23. Bartlett JG, Gorbach SL, Finegold SM. J Am Geriatr Soc 2017;65:2580-6.
12. Dickson RP, Erb-Downward JR, Huff- The bacteriology of aspiration pneumo- 36. Dennis M, Lewis S, Cranswick G,
nagle GB. Towards an ecology of the lung: nia. Am J Med 1974;56:202-7. Forbes J. FOOD: a multicentre random-
new conceptual models of pulmonary mi- 24. Cesar L, Gonzalez C, Calia FM. Bacte- ised trial evaluating feeding policies in
crobiology and pneumonia pathogenesis. riologic flora of aspiration-induced pul- patients admitted to hospital with a re-
Lancet Respir Med 2014;2:238-46. monary infections. Arch Intern Med 1975; cent stroke. Health Technol Assess 2006;
13. Dickson RP, Erb-Downward JR, Huff- 135:711-4. 10:iii-iv, ix-x, 1-120.
nagle GB. Homeostasis and its disrup- 25. Lorber B, Swenson RM. Bacteriology 37. American Geriatrics Society Ethics
tion in the lung microbiome. Am J Physiol of aspiration pneumonia: a prospective Committee and Clinical Practice and Mod-
els of Care Committee. American Geriat- Early identification of patients at risk of 65. Kadowaki M, Demura Y, Mizuno S,
rics Society feeding tubes in advanced acute lung injury: evaluation of lung injury et al. Reappraisal of clindamycin IV mono-
dementia position statement. J Am Geri- prediction score in a multicenter cohort therapy for treatment of mild-to-moder-
atr Soc 2014;62:1590-3. study. Am J Respir Crit Care Med 2011; ate aspiration pneumonia in elderly pa-
38. van der Maarel-Wierink CD, Vanob- 183:462-70. tients. Chest 2005;127:1276-82.
bergen JN, Bronkhorst EM, Schols JM, de 52. Lambert AA, Lam JO, Paik JJ, Ugarte- 66. Ott SR, Allewelt M, Lorenz J, Reim-
Baat C. Meta-analysis of dysphagia and as- Gil C, Drummond MB, Crowell TA. Risk nitz P, Lode H. Moxifloxacin vs ampicillin/
piration pneumonia in frail elders. J Dent of community-acquired pneumonia with sulbactam in aspiration pneumonia and
Res 2011;90:1398-404. outpatient proton-pump inhibitor therapy: primary lung abscess. Infection 2008;36:
39. Taylor JK, Fleming GB, Singanayagam a systematic review and meta-analysis. 23-30.
A, Hill AT, Chalmers JD. Risk factors for PLoS One 2015;10(6):e0128004. 67. Kalil AC, Metersky ML, Klompas M,
aspiration in community-acquired pneu- 53. Phillips S, Liang SS, Formaz-Preston et al. Management of adults with hospital-
monia: analysis of a hospitalized UK co- A, Stewart PA. High-risk residual gastric acquired and ventilator-associated pneu-
hort. Am J Med 2013;126:995-1001. content in fasted patients undergoing monia: 2016 clinical practice guidelines
40. Noguchi S, Yatera K, Kato T, et al. Im- gastrointestinal endoscopy: a prospective by the Infectious Diseases Society of Amer-
pact of the number of aspiration risk fac- cohort study of prevalence and predictors. ica and the American Thoracic Society.
tors on mortality and recurrence in com- Anaesth Intensive Care 2015;43:728-33. Clin Infect Dis 2016;63(5):e61-e111.
munity-onset pneumonia. Clin Interv Aging 54. Lin L, Lv L, Wang Y, Zha X, Tang F, Liu 68. Lascarrou JB, Lissonde F, Le Thuaut
2017;12:2087-94. X. The clinical features of foreign body A, et al. Antibiotic therapy in comatose
41. Perbet S, Mongardon N, Dumas F, aspiration into the lower airway in geriat- mechanically ventilated patients follow-
et al. Early-onset pneumonia after cardiac ric patients. Clin Interv Aging 2014;9: ing aspiration: differentiating pneumonia
arrest: characteristics, risk factors and in- 1613-8. from pneumonitis. Crit Care Med 2017;45:
fluence on prognosis. Am J Respir Crit 55. Miyashita N, Kawai Y, Tanaka T, et al. 1268-75.
Care Med 2011;184:1048-54. Detection failure rate of chest radiogra- 69. Momosaki R. Rehabilitative manage-
42. Johnson NJ, Carlbom DJ, Gaieski DF. phy for the identification of nursing and ment for aspiration pneumonia in elderly
Ventilator management and respiratory healthcare-associated pneumonia. J Infect patients. J Gen Fam Med 2017;18:12-5.
care after cardiac arrest: oxygenation, Chemother 2015;21:492-6. 70. Adnet F, Borron SW, Finot MA, Mi-
ventilation, infection, and injury. Chest 56. El-Solh AA, Vora H, Knight PR III, nadeo J, Baud FJ. Relation of body position
2018;153:1466-77. Porhomayon J. Diagnostic use of serum at the time of discovery with suspected
43. Rello J, Diaz E, Roque M, Vallés J. procalcitonin levels in pulmonary aspira- aspiration pneumonia in poisoned co-
Risk factors for developing pneumonia tion syndromes. Crit Care Med 2011;39: matose patients. Crit Care Med 1999;27:
within 48 hours of intubation. Am J 1251-6. 745-8.
Respir Crit Care Med 1999;159:1742-6. 57. Weiss CH, Moazed F, DiBardino D, 71. Leder SB, Suiter DM. Effect of naso-
44. Sirvent JM, Torres A, El-Ebiary M, Swaroop M, Wunderink RG. Broncho- gastric tubes on incidence of aspiration.
Castro P, de Batlle J, Bonet A. Protective alveolar lavage amylase is associated with Arch Phys Med Rehabil 2008;89:648-51.
effect of intravenously administered ce- risk factors for aspiration and predicts 72. Elke G, Felbinger TW, Heyland DK.
furoxime against nosocomial pneumo- bacterial pneumonia. Crit Care Med 2013; Gastric residual volume in critically ill pa-
nia in patients with structural coma. Am 41:765-73. tients: a dead marker or still alive? Nutr
J Respir Crit Care Med 1997;155:1729-34. 58. Samanta S, Poddar B, Azim A, Singh Clin Pract 2015;30:59-71.
45. Vallés J, Peredo R, Burgueño MJ, et al. RK, Gurjar M, Baronia AK. Significance 73. Ohkubo T, Chapman N, Neal B,
Efficacy of single-dose antibiotic against of mini bronchoalveolar lavage fluid amy- Woodward M, Omae T, Chalmers J. Effects
early-onset pneumonia in comatose pa- lase level in ventilator-associated pneu- of an angiotensin-converting enzyme
tients who are ventilated. Chest 2013;143: monia: a prospective observational study. inhibitor-based regimen on pneumonia
1219-25. Crit Care Med 2018;46:71-8. risk. Am J Respir Crit Care Med 2004;169:
46. Pinargote H, Ramos JM, Zurita A, 59. Bartlett JG. Treatment of anaerobic 1041-5.
Portilla J. Clinical features and outcomes pleuropulmonary infections. Ann Intern 74. Shinohara Y, Origasa H. Post-stroke
of aspiration pneumonia and non-aspira- Med 1975;83:376. pneumonia prevention by angiotensin-
tion pneumonia in octogenarians and 60. Finegold SM. Aspiration pneumonia. converting enzyme inhibitors: results of a
nonagenarians admitted in a General In- Rev Infect Dis 1991;13:Suppl 9:S737-S742. meta-analysis of five studies in Asians.
ternal Medicine unit. Rev Esp Quimioter 61. Levison ME, Mangura CT, Lorber B, Adv Ther 2012;29:900-12.
2015;28:310-3. et al. Clindamycin compared with peni- 75. Mylotte JM. Will maintenance of oral
47. Komiya K, Ishii H, Umeki K, et al. cillin for the treatment of anaerobic lung hygiene in nursing home residents pre-
Computed tomography findings of aspi- abscess. Ann Intern Med 1983;98:466- vent pneumonia? J Am Geriatr Soc 2018;
ration pneumonia in 53 patients. Geriatr 71. 66:590-4.
Gerontol Int 2013;13:580-5. 62. Gudiol F, Manresa F, Pallares R, et al. 76. Kaneoka A, Pisegna JM, Miloro KV,
48. Lindenauer PK, Strait KM, Grady JN, Clindamycin vs penicillin for anaerobic et al. Prevention of healthcare-associated
et al. Variation in the diagnosis of aspira- lung infections: high rate of penicillin pneumonia with oral care in individuals
tion pneumonia and association with hos- failures associated with penicillin-resis- without mechanical ventilation: a system-
pital pneumonia outcomes. Ann Am Thorac tant Bacteroides melaninogenicus. Arch atic review and meta-analysis of random-
Soc 2018;15:562-9. Intern Med 1990;150:2525-9. ized controlled trials. Infect Control Hosp
49. Warner MA, Warner ME, Weber JG. 63. Sanders CV, Hanna BJ, Lewis AC. Met- Epidemiol 2015;36:899-906.
Clinical significance of pulmonary aspi- ronidazole in the treatment of anaerobic 77. Klompas M, Speck K, Howell MD,
ration during the perioperative period. infections. Am Rev Respir Dis 1979;120: Greene LR, Berenholtz SM. Reappraisal of
Anesthesiology 1993;78:56-62. 337-43. routine oral care with chlorhexidine glu-
50. Raghavendran K, Nemzek J, Napoli- 64. Perlino CA. Metronidazole vs clinda- conate for patients receiving mechanical
tano LM, Knight PR. Aspiration-induced mycin treatment of anerobic pulmonary ventilation: systematic review and meta-
lung injury. Crit Care Med 2011;39:818-26. infection: failure of metronidazole ther- analysis. JAMA Intern Med 2014;174:751-
51. Gajic O, Dabbagh O, Park PK, et al. apy. Arch Intern Med 1981;141:1424-7. 61.
78. Higashiguchi T, Ohara H, Kamakura 79. Soutome S, Yanamoto S, Funahara M, 80. Juthani-Mehta M, Van Ness PH, Mc-
Y, et al. Efficacy of a new post-mouthwash et al. Effect of perioperative oral care on Gloin J, et al. A cluster-randomized con-
intervention (wiping plus oral nutritional prevention of postoperative pneumonia trolled trial of a multicomponent inter-
supplements) for preventing aspiration associated with esophageal cancer sur- vention protocol for pneumonia prevention
pneumonia in elderly people: a multi- gery: a multicenter case-control study among nursing home elders. Clin Infect
center, randomized, comparative trial. with propensity score matching analysis. Dis 2015;60:849-57.
Ann Nutr Metab 2017;71:253-60. Medicine (Baltimore) 2017;96(33):e7436. Copyright © 2019 Massachusetts Medical Society.